纺织学报 ›› 2019, Vol. 40 ›› Issue (11): 45-49.doi: 10.13475/j.fzxb.20181100806

• 纺织工程 • 上一篇    下一篇

基于通用字典的局部机织物纹理稳定表征

吴莹1,2(), 占竹3, 汪军3,4   

  1. 1.浙江理工大学 服装学院, 浙江 杭州 310018
    2.浙江理工大学 服装数字化技术浙江省工程实验室, 浙江 杭州 310018
    3.东华大学 纺织学院, 上海 201620
    4.东华大学 纺织面料技术教育部重点实验室, 上海 201620
  • 收稿日期:2018-11-02 修回日期:2019-03-07 出版日期:2019-11-15 发布日期:2019-11-26
  • 作者简介:吴莹(1988—),女,讲师,博士。主要研究方向为机织物纹理表征分类及其应用。E-mail: ying012688@zstu.edu.cn
  • 基金资助:
    国家自然科学基金项目(61379011);浙江理工大学科研启动基金项目(18072247-Y)

Local fabric texture stability characterization using general dictionary

WU Ying1,2(), ZHAN Zhu3, WANG Jun3,4   

  1. 1. School of Fashion Design & Engineering, Zhejiang Sci-Tech University, Hangzhou, Zhejiang 310018, China
    2. Zhejiang Province Engineering Laboratory of Clothing Digital Technology, Zhejiang Sci-Tech University, Hangzhou, Zhejiang 310018, China
    3. College of Textiles, Donghua University, Shanghai 201620, China
    4. Key Laboratory of Textile Science & Technology, Ministry of Education, Donghua University, Shanghai 201620, China
  • Received:2018-11-02 Revised:2019-03-07 Online:2019-11-15 Published:2019-11-26

摘要:

为改善固定字典表征织物纹理的效果,提出基于通用字典表征织物纹理的方法。首先,对字典学习法进行优选,分别选取8种任意正常织物样本和20种不同组织结构样本,采用优选后的字典学习法得到了1种普通通用字典、4种组织结构通用字典和1种联合组织结构通用字典。最后对通用字典的有效性进行试验验证,发现普通通用字典较固定字典能更好地适应织物纹理;在此基础上,比较了3种通用字典的性能。结果表明:在同等试验条件下,与固定字典相比,通用字典能更好地表征织物纹理;不同类型的通用字典可具有较好的通用或专用性。

关键词: 织物纹理, 字典学习, 通用字典, 织物组织结构

Abstract:

In order to improve the effect of fixed dictionary on fabric texture, a method for characterizing fabric texture based on general dictionary was proposed. Firstly, the dictionary learning method was preferred. Secondly, eight kinds of arbitrary normal fabric samples and 20 samples with different weave patterns were selected respectively, and then a general dictionary, a general dictionary of four organizational structures and a general dictionary of joint organizational structure were obtained by the preferred dictionary learning method. Finally, the validity of the general dictionary was verified by experiments. The results show that the common general dictionary can adapt to the fabric texture better than the fixed dictionary. On this basis, the performance of three general dictionaries was compared. The test results show that compared with fixed dictionaries, general dictionaries can better characterize fabric texture under the same experiment conditions; and different types of general dictionaries can have better universality or specificity.

Key words: fabric texture, dictionary learning, general dictionary, fabric weave pattern

中图分类号: 

  • TS101.9

图1

通用字典表征织物纹理的流程图"

图2

学习字典"

表1

2种字典学习算法的结果比较"

学习算法 编号 PSNR/dB SSIM 时间/s
K-SVD字典
学习法
1# 44.40 0.96 566.50
2# 30.80 0.93 563.14
MOD字典
学习法
1# 43.40 0.95 549.32
2# 30.57 0.93 538.65

图3

织物样本原始图像"

图4

通用字典"

表2

2种通用字典的量化测试结果"

字典 编号 PSNR/dB SSIM 时间/s
基于MOD算法
的通用字典
1# 39.98 0.88 89.31
2# 27.71 0.71 89.12
基于K-SVD算法
的通用字典
1# 41.53 0.92 89.36
2# 30.53 0.86 89.08

图5

2#样本在稀疏基数T值的重构图像"

图6

稀疏基数的织物纹理表征效果"

图7

2种字典的对比实验"

图8

织物样本图像"

表3

3种通用字典的对比结果"

通用
字典
PSNR/dB SSIM
2# 3# 4# 5# 2# 3# 4# 5#
组织结构 34.93 35.03 30.82 26.66 0.95 0.99 0.96 0.91
联合组
织结构
34.94 34.53 30.40 26.48 0.95 0.98 0.95 0.91
普通 33.27 32.36 29.50 26.10 0.93 0.97 0.94 0.91
[1] 薛乐, 李立轻, 汪军. 应用频谱图的机织物纹理分析[J]. 纺织学报, 2015,36(7):55-60.
XUE Le, LI Liqing, WANG Jun. Study on woven fabric tetxure using spectrum[J]. Jounrnal of Textile Research, 2015,36(7):55-60.
[2] LI P, ZHANG H, JING J, et al. Fabric defectdetection based on multi-scale wavelet transform and Gaussian mixture model method[J]. Journal of The Textile Institute, 2014,106(6):1-6.
[3] 王钢, 周建, 汪军, 等. 采用奇异值分解的机织物疵点检测算法[J]. 纺织学报, 2014,35(7):61-66.
WANG Gang, ZHOU Jian, WANG Jun, et al. Woven fabric defect detection using singular value decomposition[J]. Journal of Textile Research, 2014,35(7):61-66.
[4] 田承泰, 步红刚, 汪军, 等. 基于时间序列分形特征的织物疵点检测[J]. 纺织学报, 2010,31(5):44-47.
TIAN Chengtai, BU Honggang, WANG Jun, et al. Fabric defect detection based on fractal feature of time series[J]. Journal of Textile Research, 2010,31(5):44-47.
[5] 朱秋平. 基于稀疏编码的织物疵点检测算法研究[D]. 武汉:武汉大学, 2014:1-5.
ZHU Qiuping Research on fabric defect detection via sparse coding[D]. ZHU Qiuping Research on fabric defect detection via sparse coding[D]. Wuhan:Wuhan University, 2014:1-5.
[6] 吴莹, 汪军, 周建. 基于离散余弦变换过完备字典的机织物纹理稀疏表征[J]. 纺织学报, 2018,39(1):157-163.
WU Ying, WANG Jun, ZHOU Jian. Sparse represent- tation of woven fabric texture based on DCT over-complete dictionary[J]. Journal of Textile Research, 2018,39(1):157-163.
[7] ELAD M. Sparse and Redundant Representations from Theory to Applications in Signal and Image Process-ing[M]. New York: Springer, 2010: 227-244.
[8] ENGAN K, AASE SO, HUSOY JH. Frame based signal compression using method of optimal direc-tions (MOD) [C]// IEEE International Symposium on Circuits and Systems. Orlando: IEEE, 1999:1-4.
[9] AHARON M, ELAD M, BRUCKSTEIN A. K-SVD: an algorithm for designing overcomplete dictionaries for sparse representation[J]. IEEE Transactions on Signal Processing, 2006,54(11):4311-4322.
doi: 10.1109/TSP.2006.881199
[10] 吴莹, 汪军. 基于K-SVD学习字典的机织物纹理表征及应用[J]. 纺织学报, 2018,39(2):165-170.
WU Ying, WANG Jun. Woven fabric texture representation and application based on K-SVD dictionary[J]. Journal of Textile Research, 2018,39(2):165-170.
[1] 朱浩 丁辉 尚媛园 邵珠宏. 多纹理分级融合的织物缺陷检测算法[J]. 纺织学报, 2019, 40(06): 117-124.
[2] 李敏 赵影 张丽平 张奕 付少海. 涤纶针织物数码印花清晰度的影响因素[J]. 纺织学报, 2018, 39(05): 62-66.
[3] 吴莹 汪军. 基于K-SVD学习字典的机织物纹理表征及应用[J]. 纺织学报, 2018, 39(02): 165-170.
[4] 王凯 吴莹 周建 汪军 李立轻. 机织物密度对字典学习纹理表征的影响[J]. 纺织学报, 2017, 38(07): 142-147.
[5] 毛兆华 汪军 周建 卜佳仙 陈霞 李立轻. 应用非负字典学习的机织物瑕疵检测算法[J]. 纺织学报, 2016, 37(3): 144-149.
[6] 薛乐 李立轻 汪军. 应用频谱图的机织物纹理分析[J]. 纺织学报, 2015, 36(07): 55-60.
[7] 汪秀琛, 李晓久. 基于复合条件的织物纹理模糊识别[J]. 纺织学报, 2012, 33(8): 55-58.
[8] 刘哲;李晓久. 织物纹理均匀度的计算机评价[J]. 纺织学报, 2009, 30(12): 125-128.
[9] 徐琪. 一种支持快速响应的服装供应链补货策略[J]. 纺织学报, 2009, 30(12): 129-133.
[10] 诸葛振荣;徐利祥. 利用纹理几何约束进行织物场景模拟[J]. 纺织学报, 2007, 28(6): 55-58.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!